Air Pollution

1 Libraries

library(saqgetr)
library(tidyverse)
library(stringr)
library(lubridate)
library(gridExtra)

2 Available Data

# General data availability per site 
sites <- get_saq_sites()
sites_kiel <- sites %>% filter(str_detect(sites$site_name, "^[Kk]iel\\b"))
sites

# Detailed data availabilty 
processes <- get_saq_processes()
#unique(processes$variable_long)
processes

# NO2 Data for Kiel Theodor-Heuß-Ring since 2012.
df <- get_saq_observations(site = "desh052", variable = "no2", start = 2012, end = 2021)
df

getSeason <- function(DATES) {
    WS <- as.Date("2012-12-15", format = "%Y-%m-%d") # Winter Solstice
    SE <- as.Date("2012-3-15",  format = "%Y-%m-%d") # Spring Equinox
    SS <- as.Date("2012-6-15",  format = "%Y-%m-%d") # Summer Solstice
    FE <- as.Date("2012-9-15",  format = "%Y-%m-%d") # Fall Equinox

    # Convert dates from any year to 2012 dates
    d <- as.Date(strftime(DATES, format="2012-%m-%d"))

    ifelse (d >= WS | d < SE, "Winter",
      ifelse (d >= SE & d < SS, "Spring",
        ifelse (d >= SS & d < FE, "Summer", "Fall")))
}

3 Clean up data

• transform dates
• transform units
• remove NA records

# check for NA processes
#x <- processes %>% group_by(variable) %>% summarise(n=n_distinct(variable_long)) %>% filter(n>1)
#processes %>% filter(variable==x$variable) %>% distinct(variable, variable_long, unit)

# create all kind of dates and 
#df$unit %>% unique()

df <- get_saq_observations(site = sites_kiel$site, start = 1970, valid_only = TRUE) %>% filter(!is.na(unit)) %>%
  mutate(datetime=date,
         date=as_date(datetime),
         month_date=floor_date(datetime, "month"),
         month=month(datetime),
         month_name=month(datetime, label = TRUE),
         year=year(datetime),
         weekday=substr(weekdays(datetime),1,2),
         hour=hour(datetime),
         season=getSeason(datetime),
         value=case_when(
           str_detect(unit, "^ug") ~ value/1e3,
           str_detect(unit, "^ng") ~ value/1e6,
           str_detect(unit, "^mg") ~ value
           )
         )

translate_variable <- processes %>% filter(!is.na(variable_long), !is.na(unit)) %>% distinct(variable, variable_long) %>% mutate(variable_combined=sprintf("%s [%s]", variable_long, variable))

df <- df %>% left_join(translate_variable, by="variable")

df

4 evolution and trends of air pollution kiel.

• data from 1970 -> 2020
• only kiel
• summary=1 -> hourly data
• filtered for top 10 air pollutants over the complete timeframe
• sum of emission per air pollutant per year

# top 10 air pollutions over all time
filter_variables <- df %>% filter(summary==1) %>% group_by(variable_combined) %>% summarise(sum_value=sum(value, na.rm = TRUE)) %>% arrange(desc(sum_value)) %>% head(10)

df %>% filter(variable_combined==filter_variables$variable_combined, summary==1, year!=2021) %>% group_by(variable_combined, year) %>% summarise(sum_value=sum(value, na.rm = TRUE)) %>%
  ggplot(aes(year, sum_value, color=variable_combined)) +
  geom_line(size=1) +
  scale_color_brewer(palette = "Paired") +
  labs(x="Year", y=bquote("Chemical concentration (" ~ frac(mg,m^3) ~ ")"), colour="Air Pollutants", title="Chemical concentrations in the air over the years in Kiel") +
  theme_bw() +
  theme(plot.title = element_text(size=18, face="bold", hjust = 0.5),
        axis.text = element_text(size=16),
        axis.title = element_text(size=16),
        legend.title = element_text(size=13, face="bold"),
        legend.text = element_text(size=12)
    )

## `summarise()` has grouped output by 'variable_combined'. You can override using the `.groups` argument.

# most common summaries (summary -> 20: daily; summary -> 1: hourly)
#df %>% group_by(variable, summary) %>% count() %>% group_by(summary) %>% summarise(n_distinct_variables=length(variable)) %>% arrange(desc(n_distinct_variables))

5 seasonal, weekly, and hourly patterns of NO2 etc in germany.

• data from 1970 -> 2020
• only kiel
• summary=1 -> hourly data
• filtered for top 10 air pollutants over the complete timeframe
• timeframes: hourly/weekdays/monthly
• avg air pollution over timeframes
• normalized data to be able to compare the different air pollutants

grid.arrange(
  # p1 - hours
  df %>% filter(variable_combined==filter_variables$variable_combined, summary==1) %>% group_by(variable_combined, hour) %>% summarise(avg_value=mean(value)) %>% mutate(avg_value=scale(avg_value)) %>%
  ggplot(aes(x=hour, y=variable_combined, fill=avg_value)) +
  geom_tile() +
  theme_minimal() +
  scale_fill_distiller(palette = "Reds", direction = 1, breaks = seq(-1.5, 2, 3.5), labels = c("low", "high")) +
  labs(x="Time (h)", y="Air Pollutants", title="Hourly, weekly and monthly patterns of chemical concentrations in the air in Kiel over all years", fill="avg chemical\nconcentration") +
  theme(plot.title = element_text(size=14, face="bold", hjust = 1),
        axis.text = element_text(size=10),
        axis.title = element_text(size=10),
        legend.position = "right"
        ),
  # p2 - weekdays
  df %>% filter(variable_combined==filter_variables$variable_combined, summary==1) %>% group_by(variable_combined, weekday) %>% summarise(avg_value=mean(value))%>% mutate(avg_value=scale(avg_value)) %>%
  ggplot(aes(x=weekday, y=variable_combined, fill=avg_value)) +
  geom_tile() +
  theme_minimal() +
  scale_fill_distiller(palette = "Reds", direction = 1) +
  labs(x="Weekday", y="Air Pollutants") +
  theme(axis.text = element_text(size=10),
        axis.title = element_text(size=10),
        legend.position = "none"
    ),
    # p3 - months
  df %>% filter(variable_combined==filter_variables$variable_combined, summary==1) %>% group_by(variable_combined, month_name) %>% summarise(avg_value=mean(value))%>% mutate(avg_value=scale(avg_value)) %>%
  ggplot(aes(x=month_name, y=variable_combined, fill=avg_value)) +
  geom_tile() +
  theme_minimal() +
  scale_fill_distiller(palette = "Reds", direction = 1) +
  labs(x="Month", y="Air Pollutants") +
  theme(axis.text = element_text(size=10),
        axis.title = element_text(size=10),
        legend.position = "none"
    ),
  nrow = 3)

## `summarise()` has grouped output by 'variable_combined'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'variable_combined'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'variable_combined'. You can override using the `.groups` argument.

6 comparision of emissions across Germany to Kiel and other Cities YTD.

• data from 2020 -> 2021 filtered to YTD
• complete germany
• extracted some cities with regex
• summary=1 -> hourly data
• first plot kiel (other cities) compared to germany, which has nearly zero impact to the overall emissions
• second plot to show kiel in comparision to other cities

sites_germany <- sites %>% filter(country_iso_code=="DE") %>% distinct(site, site_name)

df_germany <- get_saq_observations(site = sites_germany$site, start = 2020, valid_only = TRUE) %>% filter(!is.na(unit)) %>%
  mutate(datetime=date,
         date=as_date(datetime),
         month_date=floor_date(datetime, "month"),
         month=month(datetime),
         month_name=month(datetime, label = TRUE),
         year=year(datetime),
         weekday=substr(weekdays(datetime),1,2),
         hour=hour(datetime),
         season=getSeason(datetime),
         value=case_when(
           str_detect(unit, "^ug") ~ value/1e3,
           str_detect(unit, "^ng") ~ value/1e6,
           str_detect(unit, "^mg") ~ value
           )
         )

translate_variable <- processes %>% filter(!is.na(variable_long), !is.na(unit)) %>% distinct(variable, variable_long) %>% mutate(variable_combined=sprintf("%s [%s]", variable_long, variable))

df_germany <- df_germany %>% left_join(translate_variable, by="variable")

#df_germany

df2 <- df_germany %>% filter(summary==1, variable_combined==filter_variables$variable_combined, date>=now() - years(1)) %>%
  inner_join(sites_germany, by="site") %>%
  mutate(city=ifelse(
  str_detect(site_name, "^[Kk]iel\\b"), "Kiel",
             ifelse(str_detect(site_name, "^[Bb]erlin\\b|^[Bb]\\b"), "Berlin",
                               ifelse(str_detect(site_name, "^[Hh]amburg\\b|^[Hh]{2}\b"), "Hamburg",
                                      ifelse(str_detect(site_name, "^[Mm][üu]e?nchen\\b"), "München",
                                             ifelse(str_detect(site_name, "^[Bb]remen\\b"), "Bremen",
                                                    ifelse(str_detect(site_name, "^[Dd]resden\\b"), "Dresden", 
                                                           ifelse(str_detect(site_name, "^[Hh]ann?over\\b"), "Hannover",
                                                                  ifelse(str_detect(site_name, "^[Kk][öo]e?ln\\b"), "Köln", "Not Defined")))))))))

grid.arrange(
  df2 %>% group_by(month_date, city) %>% summarise(sum_value=sum(value)) %>%
    ggplot(aes(x=month_date, y=sum_value, fill=city)) +
    geom_area() +
    scale_color_brewer(palette = "Paired") +
    labs(x="Date", y=bquote("Sum chemical concentration (" ~ frac(mg,m^3) ~ ")"), title="All chemical concentrations combined in Germany in comparision to Kiel (and other Cities) YTD", colour="Air Pollutants") +
    theme_bw() +
    theme(plot.title = element_text(size=14, face="bold", hjust = .1),
        axis.text = element_text(size=8),
        axis.title = element_text(size=8),
        legend.position = "right",
        legend.key.size = unit(5, 'mm')
        ),
  df2 %>% filter(city!="Not Defined") %>% group_by(month_date, city) %>% summarise(sum_value=sum(value)) %>%
    ggplot(aes(x=month_date, y=sum_value, color=city)) +
    geom_line(size=1) +
    labs(x="Date", y=bquote("Sum chemical concentration (" ~ frac(mg,m^3) ~ ")")) +
    theme_bw() +
    theme(axis.text = element_text(size=8),
        axis.title = element_text(size=8),
        legend.position = "none"
        ),
  nrow=2
)

## `summarise()` has grouped output by 'month_date'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'month_date'. You can override using the `.groups` argument.